Office Artifact: Hau Lee’s Bullwhip

After I got my tenure, I had a sabbatical. I joined Hewlett-Packard because I wanted to see how real companies operate. HP hired me full time to help solve some manufacturing problems. That year opened my eyes. That’s how I started looking at supply chain.

“Supply chain” was not a well-known term at that time. Today, it seems like everyone talks about supply chain. But not in the days when I first started.

One time, I was invited to visit Procter & Gamble with my colleagues Jin Whang [now the Jagdeep and Roshni Singh Professor of Operations, Information, and Technology, Emeritus] and Paddy Padmanabhan. The vice president of supply said, “We have a major problem because the demand for our diapers fluctuates a lot. It’s very unpredictable.”

But babies use diapers at a very stable, steady rate — there’s not much fluctuation. Yet, as we go from babies to parents, retailers, distributors, wholesalers, manufacturers, and suppliers, the order fluctuations get amplified. Why was this happening?

We observed that the same phenomenon was not just happening in diapers. It happened in apparel, it happened in semiconductors, in computers, in automobiles. Every industry has this kind of phenomenon. The key is that you have to understand it and don’t overreact, because the danger is overreacting. Sometimes you see a big demand, and you jump and produce a lot more. It turns out that is a peak, and then it will go down.

This is the bullwhip effect. Why is it like a bullwhip? When you are a cowboy and you crack the whip, you move the handle 60 degrees and the tip will move more than 360 degrees. Sometimes, a signal can move a little bit from consumer to retailer to distributor to wholesaler to manufacturer to supplier. Everyone adds something to the signal, and it results in a big surge, up and down.

The pandemic is a great illustration of this. Even the media used the term “bullwhip”: bullwhip of toilet paper, bullwhip of hand sanitizers, bullwhip of everything. We were relying on short-term data and then we produced a whole lot more. Consumers panicked, too. They didn’t realize there was ample supply coming in. Some students told me they’re still using the toilet paper that they bought in 2020.

Our paper describing the bullwhip effect was later voted one of the 10 most influential papers in the history of Management Science. Years later, I was giving a talk to executives. One was very impressed. He said, “Professor Lee, the description of the bullwhip is great, but why don’t you have a real whip to demonstrate?” A month later, he sent me a gift — a very nice box with a leather whip.

I use it in teaching. Maybe 60% of the time, I can crack the whip. I invite my students to crack the whip. They love it. I have students from 20 years ago who come back and say, “Professor Lee, I still remember your bullwhip. I don’t remember anything else. But I remember your bullwhip.” — Told to Dave Gilson